MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

What circuits benefit the most from having “character”?
MUFF WIGGLER Forum Index -> Music Tech DIY  
Author What circuits benefit the most from having “character”?
Jaytee
Conversely, what circuit benefit the least from having “character”? Why do you think so?

I was just pondering synth stuff and got to thinking hownthere are certain types of circuits where everyone clamors over capturing the exact sound of some vintage synth—VCFs are probably the canonical example.

Meanwhile, there are also circuits that nobody seems to care much about where they came from, so long as they work as expected. I’ve been thinking about sample & hold circuits all day, so that’s what came to mind. If an S&H differs much from the “ideal,” it’s no longer really a sample and hold. Nobody cares about cloning vintage sample and hold circuits, they just worry about whether it does what it says it’ll do.

There are in-between examples too. Some VCAs might color the signal in a desirable way, and so get used at the end of a signal chain, while other times you’ll want a more “transparent” VCA for utilitarian purposes.

Anyway, this is apropos of nothing in particular, I was just hoping to start an open-ended discussion and see what folks think about this stuff.
Revok
Jaytee wrote:
Nobody cares about cloning vintage sample and hold circuits, they just worry about whether it does what it says it’ll do.


I think the SSG and Source of Uncertainty have been cloned pretty heavily.
Jaytee
Revok wrote:
Jaytee wrote:
Nobody cares about cloning vintage sample and hold circuits, they just worry about whether it does what it says it’ll do.


I think the SSG and Source of Uncertainty have been cloned pretty heavily.


Alright, then what is it that makes those desirable over any other S&H?
flts
This isn't meant to be sarcasm, but it's perhaps an understatement to call SSG or SoU "sample or hold circuits". Yes, that's one part of what they do, but I think the reason why they're liked so much would rather be the whole interplay between the different parts, of which sample and hold is one.

But that does bring up a good point. I was going to say something boring on the lines of "well, anything involving audio generation or processing could benefit from its own 'character', and anything involving CV generation or processing less so".

However, the immediate counter-example is Wogglebug (which is a bunch of circuits, in turn, based on Buchla SoU) which I consider having its very definite own "character", yet what I use it mostly is for generating random control voltages.

So I don't know. Aside precision pitch generation and manipulation where you need to eg. add 1V to the input at the push of a button, I guess practically anything can be fun if it's "quirky" or "non-ideal"?
ashleym
Probably a fair number! I’m sure some have heard an owner declaring their Minimoog or Stratocaster as “a good one”. That could be true or it could be the aging has “gone well” or it’s just something people say.

I’m sure there’s mileage in an envelope altering its characteristics over the years to become faster/more linear/more exponential etc. I do thin the quirks from the classic synths can become what people define as character. You’ll hear less about a Jen SX1000 envelope and filter than you do for the Mini’s.

Waffle over. There’s potential for most areas to have character. Misshappen LFO waves, wonky envelope times, peculiar filter resonance, interesting keyboard tracking. As ever, one persons Christmas dinner is another persons poison. When does an envelope become too clicks in its quest for speed etc
Joe.
A compressor will have different character based on the design; Opto, Fet, or IC like a THAT4301.

I suppose any design (LPG etc) that can swap out the Vactrol for a faster/slower version has a tweakable character too.

The only circuits I'd say dont have character are Logic, whether it's made from diode-resistor, diode–transistor or transistor–transistor logic. Design a vacuum tube logic module, especially with a vacuum tube flipflop, and plenty of people would build it. It'd have that aesthetic character cool
flts
Joe. wrote:
The only circuits I'd say dont have character are Logic, whether it's made from diode-resistor, diode–transistor or transistor–transistor logic. Design a vacuum tube logic module, especially with a vacuum tube flipflop, and plenty of people would build it. It'd have that aesthetic character cool


People do misuse logic blocks for distortion though, eg. discrete transistor logic has a different audio rate character from CMOS hihi
slow_riot
An interesting thought is the overwhelming majority of synthesizers praised for their sonics were designed solely based on implementation of mathematics. This is certainly true of Moog, Buchla, EMS. It takes a visionary mind and true spirit to generate a completely new concept and realise it in the best possible way. These people who come after chasing "tone" and "character" are generally just offering minor alterations to these original instruments, and nothing new on an algorithmic level.

This comes up quite a lot in the euro section with people asking which is the complex osc with the most character, the answer of course is none. They are all just lesser versions of the 258, 259 etc.
papz
Revok wrote:
Jaytee wrote:
Nobody cares about cloning vintage sample and hold circuits, they just worry about whether it does what it says it’ll do.


I think the SSG and Source of Uncertainty have been cloned pretty heavily.


The Buchla random voltage generators such as the Source of Uncertainty or the 208's random (actually pseudo random) are not S&H.

Many people are wrong when it comes to what a S&H actually is and call erroneously random voltages "S&H".
A S&H is NOT a random voltage generator but a circuit with a sample input and a clock input that samples the CV present at its sample input when the clock signal comes and holds it untill a new signal comes at its clock input.
A S&H outputs random voltages when noise is present at its sample input only, which is a straightforward way to get random voltages that is widely used, hence the confusion.

It is the same S&H circuit that also keeps an analogue keyboard note CV after the key is released and until the next key is pressed, you don't want a keyboard to output random voltages, do you ?

If a LFO ramp is present at the sample input, it will output a staircase CV, not random either.
flts
papz wrote:
The Buchla random voltage generators such as the Source of Uncertainty or the 208's random (actually pseudo random) are not S&H.


Oh yeah, I didn't think that Revok might have misunderstood what a S&H is.

But the other side of the story, as I mentioned before, is that Source of Uncertainty DOES have a sample and hold block - and SSG as well. The appeal of those modules is just not specifically in the S&H part, but the fact that they produce (pseudo)random voltages that can be processed & patched in multiple ways (via S&Hs, lag generators etc.) inside the module.
KSS
Quote:
and anything involving CV generation or processing less so".

Never have understood how anyone could claim this with a straight face. (And know that the person quoted has not done so here, that name deleted for this reason)

But for those who do claim that somehow the circuits generating the CVs that control audio matter less, it just makes no sense. If audio is a rope tied to a doorknob some distance away, then CV is the person's arm and hand holding the free end and cyclicly moving it up and down. How can the accuracy and granularity of the moving arm not matter to the waveform generated? The distance, speed, direction, lag between changes each and all matter to the shape of the rope waveform created.

Would not be surprised to find that the JP4's unique EGs do have a considerable impact on the circuits they're feeding, and that the favored sound of a JP-4 to its siblings felt by many is more dependent on this than the claim at the head of this reply could support.

Believe also that the now nearly universal modern approach in commercial synths to use digital LFOs and EGs is a large part of the oft-shared sterility or too clean so-called modern sound found unsatisfying according to many posts read online about the newer synths. No, there is much important to sound living in the CV generating parts of a synth.

Character would seem to be the distance between ideal and realized results. And that is going to be true for all circuits. And all digital as well. Whether that distance is characterized as good or bad is a different question. In the S/H example of the OP, the time taken to aquire the sample, its accuracy and its decay time and curve all are some distance from the S/H ideal. Which of these and how far adds up to what would be called character. Contrary to the OP's stated position, and supported by prior replies many do want specific choices in these distances and direction away from ideal for an S/H and do go after vintage designs to acquire it.

To say this only applies with S/H would find no agreement. All circuits have character. Which benefit more or less? Isn't that the decision every person using or designing synths makes for themself? By using one over another, or choosing one circuit or element over another?

Just posted a reply in another thread about two classic MM VCFs whose difference is the buffer circuit between the VC integrators. These two filters both sound good. They also sound different. On those two points most anyone would agree. How do you refine that to say whether MM VCF's benefit more or less than say VCOs from this small circuit difference and its resulting character?

Statements from some well-respected designers show up in old threads about VCOs and their differences and make the claim that a saw is a saw is a saw. And yet, one saw does sound different than another. Because only a perfect saw compared to another perfect saw meets the saw is saw claim. And we have very few, if any, examples of perfect saws, or any other waveform. Even when we do have at least theoretical equivalence as in digital Osc's, there are still variations in how the control is mapped to these perfectly described by math or looked up tabled waves. As someone earlier pointed out, even CMOS digital logic sounds different than TTL.

I don't see how one class of circuits can be labeled as more likely to benefit from character for another reason too.

One person focuses on the effect of filtering while another focuses on waveshaping as their primary timbre tool. So the one says filter character matters more and the other says no, it's the VCO character which matters. This is not a pedantic claim. I have seen these statements from specific people here and can picture their view of which parts matter most to character when considering the sum of the posts they've made.
degeneratedsines
One of my analog S&H is drooping like hell but I like it this way, it can sometimes give an additional "modulation effect" for free hihi
I have other S&H for precise work.
Same for one of my filter which resonance is completely crazy but I never dare to adjust the dedicated trimmer on the back of the module of fear I would lose those beautiful screams.
Also quoting from RYO website about the logic modules:
Quote:
The RTL circuit design gives the modules some interesting quirks thanks to the transistors threshold windows when playing around with different amplitudes, waveforms and frequency rates of the incoming signals, making it fun and useful to explore beyond typical logic gating duties, for example as an audio waveshaper/distortion.

I never really tried it though.
KSS
degeneratedsines wrote:
Also quoting from RYO website about the logic modules:
Quote:
The RTL circuit design gives the modules some interesting quirks thanks to the transistors threshold
I never really tried it though.

Only don't assign any magic to RTL as compared to any other logic type which also has threshold windows which vary by family. CMOS at 1/3 and 2/3 supply, TTL at 0.8 and 2.4V.

One of the most character-filled ways to apply logic regardless of underlying family is to use VC comparator front ends to adjust the trip points for inputs with intent.

Or to follow the RYO idea and approach it by adjusting levels to suit the logic family.
groove
Jaytee wrote:
Conversely, what circuit benefit the least from having “character”? Why do you think so?


I don't think mults or buffered mults particularly benefit from character. Pretty much everything else could, and from there I'd say it matters what you are looking to achieve. Some people are particularly clever at abusing circuits to unexpected results. I'd say those people have a lot of character.
Revok
I haven't built the SSG or SoU (yet...) so I'm no expert on either of these.

Dave Brown's write up on the SoU is excellent and a great read:
https://modularsynthesis.com/roman/buchla_266/266sou.htm
https://modularsynthesis.com/roman/buchla266/266sou_pcb1.htm
https://modularsynthesis.com/roman/buchla266/266sou_pcb2.htm

Here's what's interesting about the sample and hold specifically. You can patch this using different circuits but wouldn't that be copying the characteristics of this S&H?

"There is a divide by two on the input pulse which creates two alternating output pulses for the two S&H's individual outputs which are bipolar. These outputs are also multiplexed together to form a single S&H output that is only positive voltages."

Another interesting aspect:
"The S&H tracks while the sample pulse is true so I shortened it to 140 uS."

I'm still trying to wrap my head around the SSG so maybe someone more familiar can help out with that one. The slewing and cycling obviously jump out to me.
dekemcculo
groove wrote:
Jaytee wrote:
Conversely, what circuit benefit the least from having “character”? Why do you think so?


I don't think mults or buffered mults particularly benefit from character.


wackel kontakt SlayerBadger! SlayerBadger! SlayerBadger!
groove
dekemcculo wrote:
groove wrote:

I don't think mults or buffered mults particularly benefit from character.


wackel kontakt SlayerBadger! SlayerBadger! SlayerBadger!


hah! There you go.
Dead Banana
snercle
Everything benefits from character, sometimes CV signals benefit even more, like the minimoog envelope generator which i believe has a few miliseconds of hold before the attack begins to descend, which makes percussive notes sound like they've been processed by a really good compressor. The S&H and LFO on the korg ps-3100 are really distinctive sounding too, they are really fragile sounding, very alive
MUFF WIGGLER Forum Index -> Music Tech DIY  
Page 1 of 1
Powered by phpBB © phpBB Group